利用核函数提高随机游走模式中污染物浓度计算的效率

Using Kernel Function to Improve Efficiency of Contaminated Concentration Calculation in Radom Walk Model

  • 摘要: 本文在拉格朗日随机游走模式中引入核函数法代替质点法进行浓度计算,并提出了对核函数的频宽进行变频宽处理,将其与对流扩散时间相关联,形成变频宽核函数。通过对固定频宽高斯核函数、固定频宽伊番科尼可夫核函数、变频宽高斯核函数和变频宽伊番科尼可夫核函数的浓度计算结果与质点法以及解析解的浓度计算结果进行比较,发现变频宽伊番科尼可夫核函数在不增加模拟粒子数的情况下能有效提高浓度计算结果的精确性。

     

    Abstract: In this paper, the kernel function method was introduced into the concentration calculation of the random walk model to replace the box method. And a change to the bandwidth of the kernel function was proposed to form a frequency bandwidth kernel function by relating it to convection diffusion time. By comparing the calculated concentration of analytical solution with the fixed bandwidth Gaussian kernel function, the fixed bandwidth Epanechnikov kernel function, the frequency bandwidth Gaussian kernel function, the frequency bandwidth Epanevhnikov kernel function and the box method, the conclusion was obtained that the frequency bandwidth Epanevhnikov kernel function can efficiently improve the accuracy of the contaminated concentration calculation without increasing simulated particles.

     

/

返回文章
返回